Gachi

Find your food people — a platonic dining-matchmaking app for Social Foodies.

Stack

• Next.js (App Router) + TypeScript + Tailwind CSS
• Supabase (Postgres + Auth) for accounts, profiles, matches, and plans

Setup

1. Create a free project at supabase.com.

2. In the Supabase dashboard, open SQL Editor → New query, paste the contents of supabase/schema.sql, and run it. This creates the tables, seeds cuisines + a few sample restaurants, and sets up row-level security.

3. In Project Settings → API, copy the Project URL and anon public key.

4. Copy .env.local.example to .env.local and fill in those two values:

   cp .env.local.example .env.local
   

5. (Optional, for faster local testing) In Authentication → Providers → Email, turn off "Confirm email" so new accounts are active immediately instead of needing an email click.

6. Install dependencies and run the dev server:

   npm install
   npm run dev
   

7. Open http://localhost:3000.

How matching works

Each person picks the cuisines they love during onboarding. On the Discover tab, everyone else is ranked by taste-match % — the overlap between your cuisine picks and theirs (shared ÷ union). Tapping Match creates a mutual match immediately (no swiping/ghosting — Gachi is low-pressure by design). From the Matches tab you can lock in a restaurant to plan a meal together.

Project structure

• src/app/page.tsx — public marketing landing page
• src/app/login, src/app/signup — auth
• src/app/onboarding — pick cuisines + basic profile info
• src/app/app/* — the signed-in product (Home, Discover, Matches, Profile)
• src/lib/data.ts — server-side data fetching from Supabase
• src/lib/matching.ts — taste-match scoring
• supabase/schema.sql — database schema, seed data, RLS policies

Next steps to consider

• Real restaurant data (an API like Google Places/Yelp instead of the seeded list)
• Photo uploads for profiles (Supabase Storage)
• In-app messaging once a match is made
• Stripe integration for the $4.99/mo subscription tier